A few quick notes on discussions the administrator team has had, since we’ve fielded a lot of questions in these particular fields. We’re posting this both for transparency and to help us limit the load of having to respond to each of you individually.
- We would like to become an official nonprofit at some point, but there is a cost associated with this, we are not lawyers, and we might need to change where/how we collect donations to do so.
- We’ve upgraded the server approximately 7 times now. We’re trying to balance fiscal responsibility with server costs. We are aware that digital ocean isn’t the cheapest server and we are trying to be conservative with estimates and give ourselves extra runtime at whatever tier we are on. We’re also hoping that the upcoming lemmy version will solve a lot of our CPU-bound issues.
We are aware of the following bugs:
- the “report created” indicator flashes in the bottom left corner for some users randomly, even non-moderator users
- sometimes, briefly or without a refresh, the username in the top corner will not be your own; as far as we can tell this is purely cosmetic and is not a security issue
- sometimes the post you’re in changes to another post for no clear reason
I know most users here are against monetization, but I think that a simple Shopify storefront with behaw merch would go a long way to help entice more people to make a monitory donation. I would love a beehaw mug, stickers, and maybe even a shirt or a hat.
Good monetization: giving people tote bags, pens, t-shirts, stickers, coffee mugs, or window clings “in exchange for a donation” like PBS does.
Bad monetization: selling the ability to downvote posts, selling avatars, selling the ability to have your post stickied, and all the other ways shitty forums sell prestigeFor only
$1/mo$5/vote, I’ll give your user the ability to downvote Comments in this thread only. Act now, time is limited!Can I pay you in the self-assessed cash value of my foot pic NFTs?
It’s going to be a long wait before the flood of PM_ME_UR_ profiles covers that.
I also really do not mind donating via Patreon or Ko-Fi or GoobGab or DinkleCoin or ScrungePay or whatever the hell people use that’s ethical now.
We do have an OpenCollective where you can donate and see our expenses: https://opencollective.com/beehaw
Though please don’t donate if you need the money, we’re doing fine financially.
The approximate yearly budget and estimated annual budget are quite different, what are those two numbers representing? Unrelated: I really appreciate what you’re making happen here, and I really look forward to being a part of it.
Good question! I don’t quite know but here’s what I think.
Estimated Annual Budget is
- All active monthly subscriptions. Multiply by 12.
- All active yearly subscriptions.
- All one-time subscriptions in the past year.
- All inactive monthly subscriptions that have contributed in the past year. https://docs.opencollective.com/help/collectives/budget#budget-calculations
Approximate yearly budget is what they we’ll spend for the year based on past transactions and trends? Note that this month will definitely be quite a bit more expensive than all previous ones (Probably something like ~500$) so this approximation is most definitely wrong.
The approximate yearly budget and estimated annual budget are quite different, what are those two numbers representing?
estimated annual we have no control over, i’d note in addition to the other reply, whereas i update approximate yearly with our hosting costs per month x12. we’ve had to upgrade a bunch this month so it’s a “best estimate of what our expenses would normally be” right now, though (and i just updated it to reflect more accurately where we’ll probably be when not needing to accommodate the big reddit surge)
I really like this idea, and I could even something that artists do on SoundCloud where an album costs $5 minimum but you can write in your own price above and beyond that initial $5.
I’d happily pick up a sticker and tack on a donation.
Merch is a great way to fund, paying a designer to create some graphics could be cost-efficient.
Thank you for all the hard work! Hopefully things will eventually slow down and less pressure will be on your shoulders. This community is more relaxed and feels like a casual hang out if anything and it’s definitely a breath of fresh air. I’m glad to be here and I’ll be sticking around for sure
I mean I have a 10gbps home connection with 1gbps backup and a server rack! No but really, I came from enterprise architecture and built some massive global pubsub networks (on the tune of 900m users/devices). Be happy to give a hand on the infra side if needed!
if you’d like to chip in please join the discord or matrix- as a general statement we are not looking to have people self host beehaw for us at this point in time
Thank you for all the tireless work. I can’t even imagine how slammed all of you must be right now.
Thanks for all the hard work. It’s seen and appreciated.
On the existing bugs/issues, are these issues with the Lemmy code in general or with the specific implementation here? Do you all need help triaging or finding root cause of issues? I am a test engineer and might be able to help out with digging into finding problems.
We will post in beehaw support when we have specific issues, we did this a few days ago and there’s already merge requests to fix it
I’ll keep a closer eye on things and maybe I can lend a hand for future problems. Glad they are getting addressed!
There is, certainly, a memory leak somewhere in the Lemmy software. @Penguincoder@beehaw.org has been trying to identify the precise location(s).
I’m sure the Lemmy developers would like help triaging and finding root cause of issues! Here’s where you can find their repos : https://github.com/LemmyNet/
They’re Lemmy-wide I believe. Have a second account on lemmy.world and I’ve seen at least the first and third bugs happen there too.
Thanks for trying to figure this out. I can’t imagine home much time this is requiring of you, and bet this isn’t what you thought you’d be doing in June.
Thanks for all the work you guys are doing. It’s really heartening to see someone trying to pick up the pieces with the mindset of not tolerating the intolerant.
Seems like a lot of instances are being hit hard today. The mod (Ruud) at lemmy.world has been posting about his server upgrades as well. Here’s his latest update: https://lemmy.world/comment/102234
Not surprising, given that today’s the day five thousand subreddits go dark. People have to get their internet fix somewhere.
Currently at 7742! (according to https://reddark.untone.uk/)
Thanks for all of the work and for the regular, transparent communication!
What are the current specs of the server y’all are using ?
Beehaw.org is run off Digital Ocean, they currently have 16 vCPUs and 32 gigs of ram. They plan on hopefully scaling back down once the whole reddit migration situation has calmed down.
That actually sounds pretty cheap. Unless digital ocean is way way way more expensive than aws?
It seems that Digital Ocean is considered relatively expensive actually.
Though I hear it has good support and it’s very easy to use. I can’t speak on how much AWS costs though, I’m unsure and don’t like supporting amazon regardless.
I’m in favor of (personally) paying more to avoid AWS also
Depends. Who wants to know and why…
I think we got it on an rpi. Thanks for the donations, we managed to buy two of them.
Lol, because I’ll be spinning up an instance this coming weekend with some QoL changes that I’d like to see, but don’t think would be merged in (route schema for posts as i dont like
{url/post/{post_id}
Appreciate all the work you do in keeping this instance running! I haven’t donated yet, but plan to do so in July, in the hopes that operating costs settle out after the the final impact of 3rd party app closure on the 30th.
Appreciate the transparency and hard work! It’s impressive really with the massive influx this place is experiencing.
I set up a recurring donation today, it’s small but I hope it helps. I’m so glad I found this place, thanks for having me. You guys are amazing, the world is better for having people like you in it!
Thanks for all the hard work. I just sent $10 your way to help support this project.
Did so earlier as well!
Nice!
I think this kind of thing will only scale long term if the user base is willing to contribute something. Even a few bucks a month would go a long way if a large enough percentage did it. Ok, I’ll get off my soapbox now. 🙂
Goofy question 'cuz I’m new here, but I noticed the LemmyNet GitHub has a dockerfile showing the backend is compiled against musl. Is there still an enormous performance difference between modern glibc malloc and musl? I get it, there’s fun in cramming image sizes down as much as possible, but in my experience the advantages of musl disintegrate pretty quickly once you get past the “interesting toy” stage of things.
I don’t know how evolved your ops platform is at this point but if you can canary in some glibc built pods/containers you might find an appreciable difference.
